CRISPR This Way: Choosing Between Delivery Systems
Keywords: CRISPR delivery, plasmid, virus, mRNA, RNP
The CRISPR/Cas9 gene editing system has revolutionized biological science. From studying gene function in cell and animal models to modifying crops for agriculture and editing disease-causing genes in humans, this approach enables a wide range of genetic modifications, including gene knockout, knockin, activation and inhibition. Regardless of the application, CRISPR components must be introduced into target cells, and various delivery formats can be used, including plasmids, RNA, ribonucleoproteins (RNPs) and viruses (Figure 1). This article focuses on the advantages and disadvantages of each CRISPR delivery system, providing key insights into which method is best suited to different experimental applications, including recommendations for transitioning into preclinical and clinical studies.
Figure 1. CRISPR/Cas9 delivery methods.
Non-viral CRISPR delivery systems
Plasmid delivery
The simplest and cheapest way to introduce foreign genetic material is direct transfection of a plasmid. A single plasmid can encode both the Cas9 endonuclease and the targeting gRNA(s), or components can be separated into two different vectors. Upon entry into the cell, transcription and translation of Cas9 and transcription of the gRNA result in the formation of the gene editing complex. Because both transcription and translation are required, this method has the slowest onset of endonuclease activity compared to the other systems, and CRISPR components are expressed for longer due to the inherent stability of DNA. This can be beneficial in scenarios where target regions are not always easily accessible due to DNA being highly compacted into higher-order chromatin structures. Natural cellular processes such as chromatin remodeling can increase the accessibility of these DNA regions over time, and thus, a slower, more prolonged process can ensure sufficient access of the CRISPR gene editing machinery to the target site. However, major drawbacks of this approach include increased off-target effects due to persistent Cas9 expression leading to cleavage at non-target sites, and the risk of insertional mutagenesis caused by the integration of plasmid DNA into the host genome.
To reduce off-target effects, the inclusion of modifications within the plasmid DNA (pDNA) should be considered during vector design. For example, Cas9 can be tagged with degradation signals or coupled to inhibitory domains to reduce activity. Another approach employs inducible systems for the expression of CRISPR components, whereby Cas9 proteins are only expressed following exposure to an external stimulus, such as light or chemical reagents. A popular example of this is the tetracycline-inducible system, which utilizes the tetracycline responsive element (TRE) promoter to drive Cas9 expression only in the presence of tetracycline or one of its analogs (e.g. doxycycline). If multiple editing experiments are required on the same cell type, stable cell lines with tetracycline-inducible Cas9 can offer minimal background editing, as well as enable long-term and large-scale editing.
Production of plasmids is relatively straightforward and amenable to up-scaling. However, during plasmid preparation, there is risk of endotoxin contamination, and some of the components required for amplification in bacteria can trigger host immunity, which poses a safety concern. To overcome this, miniaturized plasmid backbones can be utilized (e.g. MiniVec™), whereby most of the bacterial elements in the plasmid backbone are removed.
Unlike viral delivery, non-viral CRISPR delivery methods require extra assistance to cross the cell membrane. For plasmids, physical methods (e.g. electroporation and microinjection) or carrier systems (e.g. lipid nanoparticles (LNPs)) can be employed (Figure 2, Table 1). Among these, LNPs are particularly favored for clinical applications. Compared to physical methods, LNPs are much less damaging to target cells and provide a protective capsule for their cargo, enhancing stability and prolonging circulation time within the bloodstream. However, relative to electroporation and virus-based delivery, LNPs have much lower efficiency due to reliance on the endosomal pathway for entry.
Reflecting the complexities of DNA delivery and enhanced risk of off-target effects associated with plasmid delivery of CRISPR components, no CRISPR-based therapies in development utilize this delivery method. Rather, plasmid delivery is a more suitable option for research-based applications due to its simple and cost-effective nature, especially if target cells can be monitored for off-target effects.
RNA delivery
The delivery of CRISPR components as RNA (Cas9 mRNA and gRNA) is a more efficient approach compared to plasmid delivery, bypassing transcription and immediately beginning translation of Cas9 in the cytoplasm. This mode of delivery is particularly well-suited to shorter-duration experiments, as the inherent instability of RNA leads to rapid degradation within the host and, therefore, transient expression of CRISPR components. Overall, this enhances specificity, preventing prolonged Cas9 expression which can increase off-target effects and is considered safer than DNA, with no risk of insertional mutagenesis. As mRNA can be produced cell-free, this has the advantage of lower risk of contamination during production. However, the RNA manufacturing process is more expensive and technically complex compared to pDNA production, making large-scale production more challenging.
Similar to pDNA, mRNA can be delivered by microinjection, electroporation, and LNPs (Table 1). For therapeutic RNA, LNP-based delivery has proven an effective approach, with both the Pfizer and Moderna COVID-19 vaccines utilizing this strategy. The promise of future CRISPR mRNA LNP therapies is reflected in the Phase 1 clinical trials for a drug that aims to treat Transthyretin Amyloidosis (ATTR). Notably, this therapy works in vivo, whereas currently approved therapies rely on ex vivo modifications, which are more limiting and costly in time and labor.
RNP-based delivery
The highest-efficiency non-viral method of delivery for CRISPR components utilizes RNPs. There is no requirement for transcription and translation to take place, so the Cas9 protein and gRNA can immediately enter the nucleus and begin genome editing. Additionally, like RNA, there is no risk of genomic integration, and expression is transient, which minimizes off-target effects. However, the susceptibility of RNPs to degradation by proteases makes this delivery method less stable compared to plasmids and mRNA, posing challenges for handling and storage, especially in a clinical context. Additionally, the production of RNPs is more labor-intensive and expensive compared to pDNA and mRNA, and during protein production, there is an increased risk of toxic contaminants, which poses a safety risk for clinical applications.
Electroporation is the most efficient method of delivering RNPs and has been successful in ex vivo gene editing and for genome engineering of mouse embryos. However, intravenous injection of LNPs containing RNPs has proven effective for the targeted delivery of CRISPR components into muscle, brain, lungs and liver.
RNP-based CRISPR therapy has demonstrated the greatest clinical success so far, with the approval of the first CRISPR-based drug, Casgevy, to treat sickle cell anemia. This drug works by electroporation of CRISPR components as RNPs into patient cells ex vivo, after which edited cells are screened and transfused back into the patient.
Figure 2. Methods of non-viral delivery of CRISPR components.
| Delivery Method | Delivery Modality | Advantages | Disadvantages |
|---|---|---|---|
| Electroporation | DNA, mRNA or RNP | Highly efficient, works on a broad range of cell types | Damaging to cells |
| Microinjection | DNA, mRNA or RNP | Efficient on a single cell level with large cargo carrying capacity | Technically demanding and damaging to cells |
| LNPs | DNA, mRNA or RNP | History of approval by FDA | Low and variable efficiency depending on cell type |
Table 1. Comparison of non-viral methods of delivery of CRISPR components.
Viral delivery
Viral vectors are a popular option for their ease of delivery of CRISPR components into cells. This is particularly useful for cells which are difficult to transfect, including differentiated cells. Depending on the type, viruses can also be utilized for efficient in vivo transduction. Utilizing viral delivery systems does come with significant challenges, primarily due to their associated cost and complex production processes. For viral vector production, additional steps are required, whereby following the cloning of CRISPR components into a transfer vector, constructs must then be packaged into viral particles. This more labor-intensive production process also makes scaling up more difficult compared to non-viral vector production. Lentivirus, AAV, and adenovirus are the most commonly-used viral vectors for CRISPR-mediated gene editing, and each of these systems offers distinct advantages and disadvantages that can influence their suitability based on specific experimental goals (Table 2).
| Delivery Method | Efficiency | Stability | Clinical Safety | Recommended Applications |
|---|---|---|---|---|
| Lentivirus | High | Low (storage in -80℃ ~6 months) |
Moderate (WT virus insertional mutagenesis) |
In vitro and ex vivo |
| AAV | Moderate | Moderate (storage in -80℃ ~1 year) |
High | In vivo |
| Adenovirus | Moderate | Moderate (storage in -80℃ ~1 year) |
Moderate (high immunogenicity) |
In vivo |
Table 2. Comparison of viral-based CRISPR delivery methods. Major advantages are highlighted in green; major disadvantages are highlighted in red.
Lentivirus-based delivery
Lentivirus has a high gene delivery efficiency, similar to electroporation, and for this reason, it is very popular for the delivery of CRISPR components, including CRISPR libraries for large-scale screens. This delivery system is also attractive because viral cargo is inserted into the host genome for long-term expression, which can be beneficial for many overexpression experiments. However, safety concerns, particularly the risk of insertional mutagenesis, have limited its in vivo clinical use. Instead, lentiviral vectors are primarily employed for research or ex vivo modifications. This approach has led to the successful development of FDA-approved therapies, namely CAR-T for cancer treatment and Zynteglo for beta-thalassemia. For delivery of CRISPR components, lentivirus can be problematic due to off-target effects associated with continued Cas9 expression. However, this can be ameliorated using inducible expression systems or integrase-deficient lentivirus (IDLV), which reduces integration rates by ~500-fold. Although IDLV generally exhibits lower transduction efficiency compared to integration-competent lentivirus, it has demonstrated comparable efficiency in post-mitotic neurons, both in vitro and in vivo. This suggests that IDLV could be particularly valuable for future clinical applications, especially those targeting the central nervous system.
AAV-based delivery
CRISPR-based AAV therapy shows great promise, particularly for in vivo applications, due to its low immunogenicity combined with low risk of insertional mutagenesis. Indeed, the intravenous administration of AAV9 delivering CRISPR components has proven successful in the treatment of HIV in both mouse and primate models.
The main limitations of AAV as a gene delivery vector are its very restricted cargo capacity (~4.2 kb) and its lower transduction efficiency compared to lentivirus. To compensate for a low cargo capacity, a smaller Cas9 derived from Staphylococcus aureus (SaCas9) can be used, or Cas9 and gRNA can be delivered by separate vectors. Another strategy splits SpCas9 across two separate vectors containing gRNA, so that following transduction, the full Cas9 protein is assembled.
Despite some clear drawbacks, AAV is particularly advantageous in that different serotypes are available which enables targeting of specific tissues. It is even possible to generate novel capsid variants via AAV capsid evolution, whereby AAV capsid mutants are generated and screened in vitro and in vivo for improved targeting and therapeutic properties.
Although there are currently no FDA-approved AAV-based CRISPR therapies, Phase 1/2 clinical trials testing the drug EBT-101, administered intravenously for the treatment of HIV, have been completed. Additionally, several AAV-based drugs are already in use, utilizing gene overexpression, such as Zolgensma for the treatment of spinal muscular atrophy.
Adenovirus-based delivery
Adenovirus-based delivery is also generally considered safer than lentivirus, with a strong safety record demonstrated in numerous clinical trials and no risk of genomic integration. Compared to lentivirus and AAV, adenovirus has a larger carrying capacity, but its high immunogenicity can hinder gene-editing efficiency. The most commonly used serotype, human adenovirus 5 (Ad5), has restricted tropism, whereby efficient transduction is limited to cells expressing the coxsackie and adenovirus receptor (CAR). Modifications of Ad5, such as Ad5/F35, have been developed to address these issues, with chimeric adenovirus enabling infection of cells lacking CAR receptors. To reduce immunogenicity, gutless adenovirus can be used, which contains only the viral sequences essential for packaging, with the added advantage of increasing cargo capacity to 33 kb.
Transition to the clinic looks promising for many potential adenovirus-based CRISPR therapies targeting a range of diseases, including inherited disorders and cancer. Studies in mice have proven the suitability of adenovirus for in vivo treatments; however, despite FDA approval of an adenoviral gene therapy, adenovirus-based CRISPR therapeutics have yet to gain FDA approval.
Summary
Selecting the optimal system for the delivery of CRISPR components or other genetic material can be challenging due to the wide range of non-viral and viral options available. Ultimately, the best choice depends on your specific experimental goals and requires careful consideration of various factors, such as whether expression should be prolonged or transient, whether your experiments are conducted in vitro or in vivo, and if you have goals to transition to the clinic (Table 3). Additionally, designing and producing CRISPR delivery vectors can be technically demanding, time-consuming, and expensive. Fortunately, VectorBuilder has extensive experience working with a range of non-viral and viral vector systems and can support you from the initial design stages all the way through to the clinic. Recent clinical trials and the licensing of the first CRISPR drug underscore the potential for the development of highly effective CRISPR-based drugs, harnessing the power of gene delivery to cure a wide range of diseases.
| Delivery System | Efficiency | Main Delivery Techniques | Stability | Production | Cost | Clinical Safety | FDA Approved CRISPR Drug | Best Choice For |
|---|---|---|---|---|---|---|---|---|
| Plasmid | Low | Electroporation Microinjection LNPs |
High | Simple and scalable | Cheap | Low (contamination and off-target effects) |
× | Gradual prolonged gene editing in vitro |
| mRNA | Moderate | Electroporation Microinjection LNPs |
Low (degradation by RNases) |
Moderately scalable | Moderate | High (transient expression reduces off-target effects) |
× | Fast and brief gene editing in vitro or in vivo |
| RNPs | High | Electroporation Microinjection LNPs |
Low (degradation by proteases) |
Low scalability | High | High (transient expression reduces off-target effects) |
Casgevy | Fast and brief gene editing and ex vivo clinical applications |
| Viral | High (depending on viral vector used) |
Direct viral transduction | Moderate (depending on virus) |
Laborious with low scalability | Very high | Dependent on virus (immunogenicity and risk of off-target effects) |
× | For difficult to transfect cells and clinical applications |
Table 3. Comparison of different modalities for delivery of CRISPR components. Major advantages are highlighted in green and major disadvantages are highlighted in red.
Sources
Yip BH. Recent Advances in CRISPR/Cas9 Delivery Strategies. Biomolecules. 2020 May 30;10(6):839.
National Library of Medicine. National Centre for Biotechnology Information. ClinicalTrials.gov. https://clinicaltrials.gov/search?intr=crispr